Transcript Mark Thomas - 1000 Lives Plus

29th September 2009
Patient Identification (Part 2)
Presenters: Dafydd Thomas & Mark Thomas
NHS Reality
NHS Reality
NHS Reality - MITU
Mrs Jones & Utopia
Patient
Gold Standard
Record
PAS
3 records
Radiology
2 records
Pathology
9 records
Signpost to other systems
Pharmacy
3 records
A&E
4 records
GP
1 record
The Solution
“Every patient must be uniquely identified
in an unambiguous manner.
This identification must be tied to all
requests, medications, procedures,
devices etc. applied to the patient.”
Australian Commission on Safety and Quality in Healthcare
(2008). Technology Solutions to Patient Misidentification. Report
of Review.
NPSA Recommendations
• Use the NHS number as the national patient
identifier on:
–
–
–
–
Patient correspondence
Patient notes
Patient wristbands
Electronic records
• Educate patients on its use
• Primary Care organisations to inform new
patients of their number
NHS Number
First name
Last name
DOB
Sex
NPSA Standard
First name
Last name
DOB
Sex
NHS Number
Baby
Thomas
21 June 2007
Male
N4Babies
All
>One
All
50%
The Times Monday 28 September 2009
To err is human but to really screw things up takes a computer
Safety
System errors
Design
Technology
Human
Aviation
F1
Tescos
Error/Violation
The Requirement
• Identify a patient unambiguously using
‘Gold Standard’ demographic
• Use identification for all patient related
activity
• Link electronic records held in different
systems, accurately & unambiguously
Prerequisite to the development of
patient centric record e.g. Welsh
Clinical Portal
IHC Projects
• Duplicate Records Project
–£5.5m invested from 2004 – 2008
–Reduction in duplicate records
–Reduction in missing NHS
numbers
0%
14%
PAS
Pathology
8%
N
ov
-0
Fe 4
b0
M 5
ay
-0
5
A
ug
-0
5
N
ov
-0
Fe 5
b0
M 6
ay
-0
6
A
ug
-0
6
N
ov
-0
Fe 6
b0
M 7
ay
-0
7
A
ug
-0
7
N
ov
-0
Fe 7
b08
Mar-08
Jan-08
Nov-07
Sep-07
10%
Jul-07
12%
May-07
Mar-07
Jan-07
Nov-06
Sep-06
Jul-06
May-06
Mar-06
Jan-06
Nov-05
Sep-05
Jul-05
May-05
Mar-05
Jan-05
Nov-04
Duplicate Records Results
Duplicate Records - All Wales Averages
Missing NHS Number - All Wales Average
80%
70%
Radiology
60%
50%
PAS
40%
Pathology
6%
30%
4%
20%
2%
10%
Radiology
0%
IHC Projects
Bar code Labelling Project
– Barcode label printers supplied to
GP surgeries
– Used for Pathology test requests
– Correct, legible patient information
on request forms
– Patient identifiers scanned into
hospital systems – eliminating
manual entry
Labelling Project Results
Time taken to complete request forms and 2 sample labels
6
Minutes
5
4
Pre-implementation
3
Post-implementation
2
1
0
1
2
3
4
5
6
7
8
9
10 11 12 13 14 15 16 17 18 Av
Practice
Average Results
18
16
14
12
10
8
6
4
2
0
Pre-implementation
Post-implementation
Samples per
week (%)
Length of time to
complete form
and 2 samples
(min)
Clinics overrunning (min)
Question
No. of contacts
Time spent
with lab chasing
discussing
results per week queries with Lab
per week (min)
Background
Separate departmental
systems
+
Separate manual (error
prone) registration
processes
+
Trust Merger
=
=
Multiple records
and identifiers for
same patient
(Multiple records
and identifiers
for same
patient)2
Merger Consequences - ABM
1 Patient with records potentially in:
• 2 PAS
• 2 Pathology
• 2 Radiology
• 2 Theatres
• 4 Pharmacy Systems...
Scale of the Problem - ABMU
• 2 PAS Duplicate & Potential Duplicates:
– iPM:
•
•
170,000 duplicates
40,000 potential duplicates
– PIMS:
•
5,000 duplicates
– Across both PAS’s:
•
•
100,000 duplicates
50,000 potential duplicates
1 Person resolving
duplicates:
• 3.6 years to resolve
potential duplicates
• 54 years to
amalgamate case
notes of duplicate
records
Impact of the Problem
• Clinicians presented with long list of
results for patients that could be the
same:
– 2000 David Jones
– 25% of Stephen Davies share date of birth
with another Stephen Davies
– 15% of records with surname Jones,
Davies, Williams, Thomas or Evans
BBTS into Google
• British Blood Transfusion Society
BBTS into Google
• Big Bad Toy Store
• Bethesda by the Sea (Chapel - USA)
• British Blood Transfusion Society
Impact of the Problem
• Patient information missing (or incorrect)
when making key clinical decisions
• Unnecessary tests re-ordered for patients as
results contained in other unknown records
Identification across Wales
• Identify patients across Wales
–
–
–
–
Regional/National Pathology services
Patients moving across boundaries
(Welsh) Clinical portal
Local identifiers not enough – may conflict with
other organisations’ identifiers
• The NHS number is only part of the solution
NHS number
• Unique number for every person registered
with a GP practice in England and Wales
• Welsh Demographic Service (WDS) is the
master store for Wales – fed from GP
systems
• Use recommended by NPSA from Sept 09
NHS number
• Issues with relying solely on the NHS
Number:
– Manually input into disparate clinical
systems
– 814 pathology & radiology results where
same hospital number has >1 NHS number
– Not always available - 50,000 records
without NHS no. in results reporting
database
– Blood Gas Scenario
Requirements – a
reminder
• Link patient based information across
different departments/organisations
(including merged
organisations)/Primary Care
• Provide ‘Gold Standard’ demographic
record in order to:
– Safely & consistently present all relevant
information to clinicians involved in care
Creating the Gold
Standard
• Patient search:
– Consistent process
– Staff training/awareness
• Registration – obtain NHS no. from WDS ‘in
real time”
• Batch trace with WDS to ensure full NHS no.
completeness
• Utilise automatic address searching tool
Creating the Gold Standard #2
• Amend demographics at PAS level
– Shared to other systems
– Up to date demographic available to all
• No manual entry of numbers
– Bar coded labels
– Electronic test requesting
eMPI – What it does
• Identifies and automatically links duplicate
records
• Identifies potential duplicates
• Enables potential duplicates to be managed in a
live environment – protecting the Gold Standard
• Signposts to records in other systems
• Shares Gold Standard demographics with other
systems
• Enables safer workflow between departments
Welsh eMPI
• IHC procuring one common eMPI
• Will be implemented 6 times in LHBs – same
• Prerequisite to safely identifying patients
across the PAS, other systems and service
• Powers local systems safely
• Powers the WCP - which pulls all the patients
data based on their ID from the different
systems into a common view of the record
• Contract signed November 09
eMPI
• This is only the
“End of the beginning”
The Bro Morgannwg
MPI
• MPI linked to PAS and clinical systems
• Matching algorithm determines definite and potential
duplicate records
• Results:
– 45,000 duplicate records reduced to 5,000 duplicates
– 40,000 PAS records (& case notes) amalgamated/volumes
linked
– 7 potential duplicates (4 definites) identified and resolved
each day
1 patient, 1 record
• Gold standard maintained
Daily Duplicate
Registrations
Average 4 per day
Key Principles
• New & consistent:
– Standards need to be in place for patient ID
across all departmental systems
– Application of these standards requires
significant changes in departmental workflow
– Adherence to these standards needs to be
proactively monitored and assured
Key Principles #2
• Ongoing resource to be allocated
– Reallocation of resource needed to provide
sustainable robust patient identification processes
Safety and Financial Reasons
• Information must be in real time to maintain gold
standard and support operational processes
The ‘How to Guide’
•
Reducing harm as a result of patient
identification errors
– Achieving the ‘Gold Standard’ in patient
identification
Driver Diagram
1. Support front-line staff to reduce the
opportunity for patient identification errors
2. Develop ‘back room’ functions to reduce
the risk of patient identification errors and
provide clinical staff with a more complete
record of care
Page 2
Process Measures
• Compliance to searching process
• Proportion of patients with standardised
bar-coded wristbands
• Number of new duplicates created per day
Page 3
Outcome Measures
• Number of cases of harm due to patient
misidentification errors
• Number of cases of harm where critical
information is missed as patient
information is unlinked across systems
Page 3
Driver 1
•
•
Support front-line staff to reduce the opportunity for
patient identification errors
6 interventions:
1.
2.
3.
4.
5.
6.
Page 7
Redesign patient searching process
Redesign patient registration process
Redesign birth & death notification processes
Promote the use of NHS Number as unique patient identifier
Identify all patients using standardised bar-coded wristbands
Use NHS Number when requesting tests and other services
Gold Standard Search &
Registration Process
Search for patient on
system to determine
whether a previous
registration exists
Barcode to be used
whenever possible
First Search
Enter the NHS
Number/Hospital
Number
Only one patient
should be
displayed
Evidence - 1st Appointments
Referral Registration hit rate
98%
Second Search
Enter patients
initials & DOB
Patients are listed
in alphabetical
order by surname
Third Search
Enter Date of Birth
Patients are listed
in date of birth
order
Fourth Search
Enter the
Postcode
Do All Key Demographic
Fields Match
- Patient
- Referral/Request
Yes
Use this record
Patients are listed
in alphabetical
order by surname
No
Yes
Fifth Search
Enter patient’s
Surname &
Forename(s)
Sixth Search
Soundex Search –
similar spellings
Page 9
Patients are listed
in alphabetical
order by surname
Brings back
patients with
similar names /
spellings
Patient Script
Verify and repeat back
the details:DOB
Surname & Forenames
and correct spellings
Ask if they have
attended the hospital
Patient
Match
Confirm details with:1. Patient ( If present)
2. WDS
3. GP
Patient cannot be
found
Amend as
appropriate
Register patient
Use this record
Driver 2
•
•
Develop ‘back room’ functions to reduce the risk of patient
identification errors and provide clinical staff with a more
complete record of care
7 interventions:
1.
2.
3.
4.
5.
6.
7.
Page 15
Implement an Enterprise Master Patient Index (EMPI)
Identify and resolve potential duplicate and duplicate patient records
Amalgamate duplicate records
Link disparate systems
Batch trace NHS Number for existing records
Provide a patient-centric view of information
Maintain ‘Gold Standard’ in patient identification
Patient
MPI
PAS
3 records
Page 16
Radiology
2 records
Pathology
9 records
Gold Standard
Record
Signpost to other systems
Pharmacy
3 records
A&E
4 records
Welsh
Demographic
Service
GP
1 record
Appendix 1
• Achieving change in practice
–
–
–
–
–
–
Getting started
Use the Model for Improvement
Forming the Team
Setting Aims
Barriers that may be encountered
Measurement
Page 19
Conclusion
• Organisations need to review their administrative
processes and patient identification policies to minimise
risk of duplicate records
• An Enterprise Master Patient Index is essential to
managing patient records safely
• A successful project has already been completed in
Wales, so it is possible for other organisations to
implement similar solutions
• The 1000Lives Campaign methodology can assist
organisational teams implement their local solutions
successfully